Science of the Total Environment 657 (2019) 1343–1356

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Review

The state of desalination and brine production: A global outlook

Edward Jones a,b, Manzoor Qadir a,⁎,MichelleT.H.vanVlietb, Vladimir Smakhtin a, Seong-mu Kang a,c a United Nations University: Institute for Water, Environment and Health (UNU-INWEH), Canada b Water Systems and Global Change, Wageningen University, the Netherlands c Gwangju Institute of Science and Technology (GIST), South Korea

HIGHLIGHTS GRAPHICAL ABSTRACT

• Unconventional water resources are key to support SDG 6 achievement. • Desalinated water production is 95.37 million m3/day. • Brine production and energy consump- tion are key barriers to desalination ex- pansion. • Brine production is 141.5 million m3/day, 50% greater than previous estimates. • Innovation and developments in brine management and disposal options are required.

article info abstract

Article history: Rising water demands and diminishing water supplies are exacerbating water scarcity in most world regions. Received 31 August 2018 Conventional approaches relying on rainfall and river runoff in water scarce areas are no longer sufficient to Received in revised form 5 December 2018 meet human demands. Unconventional water resources, such as desalinated water, are expected to play a key Accepted 5 December 2018 role in narrowing the water demand-supply gap. Our synthesis of desalination data suggests that there are Available online 07 December 2018 15,906 operational desalination plants producing around 95 million m3/day of desalinated water for human Editor: Ashantha Goonetilleke use, of which 48% is produced in the Middle East and North Africa region. A major challenge associated with de- salination technologies is the production of a typically hypersaline concentrate (termed ‘brine’) discharge that re- Keywords: quires disposal, which is both costly and associated with negative environmental impacts. Our estimates reveal Recovery ratio brine production to be around 142 million m3/day, approximately 50% greater than previous quantifications. Feedwater type Brine production in Saudi Arabia, UAE, Kuwait and Qatar accounts for 55% of the total global share. Improved Desalination technology brine management strategies are required to limit the negative environmental impacts and reduce the economic Product water cost of disposal, thereby stimulating further developments in desalination facilities to safeguard water supplies Concentrate stream for current and future generations. © 2018 Elsevier B.V. All rights reserved.

⁎ Corresponding author. E-mail address: [email protected] (M. Qadir).

https://doi.org/10.1016/j.scitotenv.2018.12.076 0048-9697/© 2018 Elsevier B.V. All rights reserved. 1344 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356

Contents

1. Introduction...... 1344 2. Methodology...... 1345 2.1. Globalstatusofdesalination:researchandpractice...... 1345 2.1.1. Desalinationinresearch...... 1345 2.1.2. Desalinationinpractice...... 1345 2.2. Brineproduction...... 1345 3. Results...... 1346 3.1. Researchtrendsindesalination...... 1346 3.2. Globalstateofdesalination...... 1346 3.3. Brineproduction...... 1349 4. Discussion...... 1352 5. Conclusions&outlook...... 1354 Acknowledgements...... 1355 AppendixA. Supplementaryinformation...... 1355 References...... 1355

1. Introduction advances, coupled with rising costs and the diminishing supply and se- curity of “conventional” water resources, have made desalination a cost- Rising water demands associated with population growth, increased competitive and attractive water resources management option around water consumption per capita and economic growth, coupled with the globe (Ghaffour et al., 2013; Sood and Smakhtin, 2014; Caldera and diminishing water supplies due to climate change and contamination, Breyer, 2017; Darre and Toor, 2018). Nowadays, an estimated 15,906 are exacerbating water scarcity in most world regions (Richter et al., desalination plants are currently operational, located in 177 countries 2013; Djuma et al., 2016; Damania et al., 2017). Recent estimates suggest and territories across all major world regions. that 40% of the global population faces severe water scarcity, rising to 60% Realising the vast potential of desalinated water remains a challenge by 2025 (Schewe et al., 2014). Furthermore, 66% of the global population due to specific barriers, predominantly associated with the relatively (4 billion) currently lives in conditions of severe water scarcity for at least high economic costs and a variety of environmental concerns (e.g. one month per year (Mekonnen and Hoekstra, 2016). These statistics Einav et al., 2002; Roberts et al., 2010; Richter et al., 2013; Darre and demonstrate that “conventional” sources of water such as rainfall, snow- Toor, 2018). Continued improvements in membrane technologies, en- melt and river runoff captured in lakes, rivers, and aquifers are no longer ergy recovery systems and coupling desalination plants with renewable sufficient to meet human demands in water-scarce areas. This is in direct energy sources provide opportunities for reducing the economic costs of conflict with Sustainable Development Goal (SDG) 6, aimed at ensuring desalination (Elimelech and Phillip, 2011; Pinto and Marques, 2017; the availability of clean water for current and future generations. Darre and Toor, 2018), whilst trends towards stricter environmental Water-scarce countries and communities need a radical re-think of guidelines and permitting factors may cause the falling trend in desalina- water resource planning and management that includes the creative ex- tion costs to slow, level off or reverse (Pinto and Marques, 2017). Regard- ploitation of a growing set of viable but unconventional water resources less, continued reductions in the economic costs of desalination will be for sector water uses, livelihoods, ecosystems, climate change adapta- required for desalination to be considered a viable option for addressing tion, and sustainable development (Qadir, 2018). Whilst water demand SDG 6 in low income countries. Detailed evaluations of the challenges mitigation approaches such as water conservation and improved effi- and opportunities associated with the economics of desalination are pro- ciencies can somewhat close the water demand and supply gap, these vided by Ghaffour et al. (2013) and Pinto and Marques (2017). approaches must be combined with supply enhancement strategies in The safe disposal of effluent produced in the desalination process order to combat water scarcity (Gude, 2017). Such water resources con- remains a particular concern and a major technical and economic servation and supply enhancement strategies are already practiced in challenge (Roberts et al., 2010). The desalination process separates some water-scarce areas. However, expansion is required, particularly intake water into two different streams – a freshwater stream (product in areas where water scarcity and water quality deterioration is intensi- water) and a concentrate waste stream (Wenten et al., 2017). The salin- fying (van Vliet et al., 2017; Jones and van Vliet, 2018). ity of the concentrate stream depends on the salinity of the feedwater. Among the water supply enhancement options, desalination of sea- As the vast majority of concentrate is produced from saline water water and highly brackish water has received the most consideration (N95% from SW and BW sources), the term ‘brine’ is used throughout and is increasingly seen as a viable option to meet primarily domestic this paper. However, it should be noted that desalination plants operat- and municipal needs. Desalination is the process of removing salts ing with low saline feedwater types (e.g. RW, FW) produce concentrate from water to produce water that meets the quality (salinity) require- with a lower salinity than typically associated with the term ‘brine’. ments of different human uses (Darre and Toor, 2018). Seawater desa- A desalination plant water recovery ratio (RR), defined as the volu- lination can extend water supplies beyond what is available from the metric processing efficiency of the purification process (Harvey, 2008), in- hydrological cycle, providing an “unlimited”, climate-independent and dicates the proportion of intake water that is converted into high quality steady supply of high-quality water (Elimelech and Phillip, 2011). (low salinity) water for sectoral use. The remaining water (calculated as Brackish surface and groundwater desalination offers reductions in the (1 − RR)) is the proportion of intake water being converted into a salinity levels of existing terrestrial freshwater resources below sectoral waste (brine) stream, which requires management. For example, a desa- thresholds (Gude, 2017). lination plant operating with a recovery ratio of 0.4 means that 40% of in- The uptake of desalination has been substantial, but limited predom- take water is converted into product water, and by extension 60% of inantly to high income countries (e.g. Saudi Arabia, UAE, Kuwait) and intake water is converted into brine. The RR of a desalination plant is de- small island nations (e.g. Malta, Cyprus) with highly limited ‘conven- pendent on and controlled by a number of factors (Xu et al., 2013). Differ- tional’ water resources (e.g. rainfall, snowmelt). However, reductions ent desalination technologies are associated with variations in RR, with in the economic cost of desalination associated with technological membrane technologies typically associated with a much higher RR E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 1345 than possible with thermal technologies (Xu et al., 2013). The feedwater Saharan Africa; and 8) Western Europe. Country data was used to assign quality is also important, with it being much more difficult (and expen- each desalination plant to one of four economic levels based on the 2018 sive) to operate desalination plants at a high level of water recovery World Bank Income groups, whereby GNI per capita ($) is estimated when the feedwater salinity is high (Harvey, 2008). using the World Bank Atlas method. Countries are assigned to one of With the aim of providing a global assessment of the research and four economic classifications: 1) High income (N$12,056 GNI per practice around desalination, the objectives of this study are to: capita); 2) Upper middle income ($3896 to $12,055); 3) Lower middle (1) share an insight into the historical development of desalination; income ($966 to $3895); and 4) Low income (b$995). (2) provide a state-of-the-art outlook on the status of desalination, consid- The sector (or ‘customer type’) for each desalination plant was sep- ering the number of desalination facilities and their associated treatment arated into six categories: 1) Municipal (including tourist drinking capacity with regards to aspects such as geographical distribution, desali- water facilities); 2) Industry; 3) Power stations; 4) Irrigation; 5) Mili- nation technologies, feedwater types and water uses; and (3) assess brine tary; and 6) Other. ‘Other’ comprises uses of Demonstration, Process production from desalination facilities and the management implications and Water Injection, which are not considered separately as they ac- of the produced brine. This study therefore seeks to update the literature count for b0.2% of total desalinated water use. on the state of desalination in both research and practice, which is out- Feedwater type is separated into six categories in DesalData (2018) dated. Furthermore, this study makes the first comprehensive quantifica- expressed in ppm Total Dissolved Solids (TDS): 1) Seawater (SW) tion of the volume of brine produced by desalination facilities, employing [20,000–50,000 ppm TDS]; 2) Brackish water (BW) [3000–20,000 ppm a novel methodology that considers the efficiency of desalination plants TDS]; 3) River water (RW) [500–3000 ppm TDS]; 4) Pure water (PW) based on both their operating technology and the feedwater type. [b500 ppm TDS]; 5) Brine (BR) [N50,000 ppm TDS]; and 6) Wastewater (WW). Despite having a typically high base quality (low salinity), desali- 2. Methodology nation of RW is practiced for a range of different sectoral uses (e.g. drink- ing water, irrigation) to reduce water salinity below specific sectoral 2.1. Global status of desalination: research and practice thresholds. PW as a feedwater source is typically used for industrial appli- cations which require very high quality (low salinity) water, such as the 2.1.1. Desalination in research pharmaceutical and food production industries. A bibliometric analysis was conducted to evaluate the major re- Desalination technology was separated into seven categories: 1) Re- search trends in the field of desalination. The Science Citation Index Ex- verse Osmosis (RO); 2) Multi-Stage Flash (MSF); 3) Multi-Effect Distil- panded (SCI-EXPANDED) from the Web of Science Core collection was lation (MED); 4) Nanofiltration (NF); 5) Electrodialysis/Electrodialysis used for the time period 1980 to 2018. This study firstly categorises de- Reversal (ED); (6) Electrodeionization (EDI); and 7) Other. ‘Other’ in- salination publications based on major research theme (‘technology’, cluded a variety of technologies such as 1) Forward Osmosis (FO); 2) Hy- ‘environment’, ‘economic and energy’ and ‘social interests’). Subse- brid (HYB); 3) Membrane distillation (MD); 4) Vapour compression quently, considering the ‘technology’ category, trends in research on (VP); and 5) Unknown. As the technologies grouped together under specific technologies (‘Reverse Osmosis’, ‘Multi-Effect Distillation’, the ‘Other’ category contribute a total of b1% of the total desalinated ‘Multi-Stage Flash’, ‘Electrodialysis’, ‘Emerging’ and ‘Other’) were exam- water produced, these technologies were not considered individually. ined. ‘Emerging’ refers to technologies largely in the R&D phase (For- ward Osmosis, Membrane Distillation and Nanofiltration) whereas 2.2. Brine production older, less prevalent technologies were categorised as ‘other’ (Humidifi- cation-Dehumidification, Solar Stills and Vapour Compression). The The volume of brine produced was determined at each individual precise methodology adopted for the bibliometric study is presented (operational) desalination plant using three factors contained in in the Supplementary material. DesalData (2018) - feedwater type, desalination technology and treat- ment capacity (m3/day). We consider the water recovery ratios associ- 2.1.2. Desalination in practice ated with different feedwater-desalination technology combinations A global database containing information on approximately 20,000 and calculate the brine production based on this recovery ratio and desalination plants (version of 2018) was obtained from Global Water the plant capacity using Eq. (1). Intelligence (GWI) (https://www.desaldata.com). The database con- Qd tains information on the plant status, operational year, plant capacity, Qb ¼ ðÞ1−RR ð1Þ geographic location (region, country, coordinates), customer type, desa- RR lination technology and feedwater type of each individual desalination whereby Qb is the volume of brine produced (m3/day); Qd is the desa- plant. The precise geographic location of each desalination plant was lination plant treatment capacity (m3/day) and; RR is the recovery ratio. plotted in ArcGIS using latitude and longitude data. The rest of the In total, 41 different feedwater type and desalination technology data was tabulated using pivot tables in Microsoft Excel to assess statis- combinations are currently operational. The recovery ratio associated tics of multiple desalination plants per region, technology and other cat- with each of these feedwater-technology combinations was determined egories. Desalination data (number and capacity of plants) was using two methods. Firstly, a literature study was conducted in order to subsequently analysed at the global, regional and national scale. The identify values of recovery ratios (or % water efficiency) for different specifics within each category by which the global state of desalination technologies and feedwater types reported in existing studies. When was analysed are as follows. recovery ratios were expressed as a range, the midpoint was used. In Plant status was categorised as either 1) Online; 2) Presumed online; total, 89 recovery ratios were found in the literature across a range of 3) Construction; 4) Presumed offline; or 5) Offline. In this study, desali- feedwater-technology combinations. Secondly, influent and effluent nation plants were considered ‘Operational’ if they were classified as ei- salinity data from individual desalination plants operating with ther ‘Online’, ‘Presumed online’ or ‘Construction’. Operational year membrane technologies was used to estimate recovery ratios using refers to the year in which the desalination plant opened, assigned Eqs. (2) and (3) (Bashitialshaaer et al., 2009). unanimously as 2020 for all plants currently in construction. Plant desa- lination capacity, or the volume of high quality product water produced Sf 3 Sb ¼ ð2Þ for human use, is provided in m /day for each desalination plant. 1−RR Eight geographic regions were identified:1)EastAsia&Pacific; 2) Eastern Europe & Central Asia; 3) Latin America & Caribbean; 4) Mid- Sf RR ¼ 1− ð3Þ dle East & North Africa; 5) North America; 6) Southern Asia; 7) Sub- Sb 1346 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 whereby Sb is the brine salinity and Sf is the feedwater salinity, with such as river regulation (e.g. dam buildings) and water transfers (March both salinities expressed in the same units (e.g. mS/cm for EC, mg/l for et al., 2014), which may in part explain the lack of publications. Further- TDS). more, desalination operations are not typically associated with the We obtained 30 additional recovery ratios using this method, which gender issues and community-based factors associated with other un- were combined with recovery ratios identified in the literature to pro- conventional water resources, such as fog water harvesting (Qadir duce 119 records. From this, average recovery ratios could be identified et al., 2018; Lucier and Qadir, 2018). However, desalinated operations for 18 of the 41 technology-feedwater. Whilst this coverage might are associated with some important (and under-researched) policy- seem low, desalination-technology combinations are not all equally related aspects, such as the lack of specific water standards for desali- prevalent in terms of number of plants and desalination capacity. nated water for both the municipal (Chen et al., 2015) and agricultural These 14 combinations account for N80% of the total desalinated water sectors (Martinez-Alvarez et al., 2016). As desalination continues to be- produced globally, with the top three combinations (seawater (SW)- come a more prevalent water resources management technology in the RO, brackish water (BW)-RO and SW-MSF) accounting for 70% of the future, the number of publications across all categories, and especially produced desalinated water alone. In order to determine recovery ratios environmental and socio-political themes, is expected to increase for the remaining feedwater-technology combinations, a number of as- rapidly. sumptions and estimations were made (Table 1). Publications addressing technological aspects have dominated the Latitude and longitude data was used to calculate the distance of research history of desalination (Fig. 1). Fig. 2 further explores this each desalination plant from the nearest coastline using the Spatial An- trend by categorising ‘technological’ publications by specific technol- alyst tool in ArcGIS. Combined with the estimated brine production for ogy. RO is the most researched technology throughout the entire time each desalination plant, we calculated the volume of brine produced period, with the number of publications approximately doubling each at different distances from the coastline to consider the implications five-year period. Research into ‘emerging’ technologies (FO, MD and for brine management. NF) is increasing at the most rapid pace with increasing recognisation of their potential advantages over existing commercial technologies. 3. Results These include factors such as operating at higher water recovery ratios and requiring less and/or sustainable energy (Subramani and Jacangelo, 3.1. Research trends in desalination 2015). Thermal technologies (MED and MSF), despite accounting for a significant share in the amount of desalinated water produced, have re- Trends in the research history of desalination are displayed in Fig. 1. ceived comparatively little attention in recent literature. Whilst publica- Approximately 16,500 publications were found to have been produced tions addressing MSF and MED accounted for a significant proportion of on the topic of desalination since 1980. Research in desalination has research in the 1980s and 1990s, they are now the overall least grown exponentially, with the total number of publications approxi- researched technologies. Concerns over the energy costs, efficiency mately doubling with each five-year period (e.g. ~5000 in 2010 to and environmental impacts of thermal processes, and the development ~11,000 in 2015). The large majority of publications focus on technolog- of more efficient membrane technologies and techniques (particularly ical aspects of desalination (e.g. 75% in 2005). As such, desalination lit- RO), likely explain this trend. erature focusing on technological aspects has driven the overall trend in desalination research. Whilst the proportion of desalination literature 3.2. Global state of desalination covering technological aspects is still high (72%), there has been an emergence of literature covering alternative aspects of desalination, There are 15,906 operational desalination plants with a total particularly related to economics and energy and environmental con- desalination capacity of approximately 95.37 million m3/day cerns. The number of publications considering economic aspects of de- (34.81 billion m3/year), constituting 81% and 93% of the total num- salination has increased dramatically in recent decades, from b400 in ber and capacity of desalination plants ever built respectively 2000 to N5000 in 2018. Historically, the environmental impacts of desa- (Fig. 3a). Early desalination plants predominantly utilised thermal lination were severely neglected, with just 118 publications before technologies, located in oil-rich but water scarce regions, espe- 2000. However, literature published in this category is now increasing cially in the Middle East. For example, prior to the 1980s, 84% of at the fastest rate, with an additional ~2000 publications since 2000. all global desalinated water was being produced by the two major The number of publications addressing socio-political aspects of desali- thermal technologies (MSF, MED). The rise in the use of membrane nation is relatively low. Desalination is not typically associated with so- technologies post-1980, in particular RO, gradually shifted the dominance cial opposition and conflict associated with other water supply schemes away from thermal technologies. In 2000, the volumes of desalinated water produced by thermal technologies (dominated by MSF) and RO were approximately equal at 11.6 million m3/day and 11.4 m3/day re- Table 1 spectively, together accounting for 93% of the total volume of desalinated Assumptions and estimations used determining the recovery ratios of feedwater-technol- water produced (Fig. 3b). Since 2000, both the number and capacity of RO ogy combinations used in operational desalination plants. plants has risen exponentially, whilst thermal technologies have only ex- Assumption perienced marginal increases (Fig. 3b). The current production of desali- 1 When brackish water (BW) recovery is known, the water recovery ratio of nated water from reverse osmosis now stands at 65.5 million m3/day, brine (BR) (TDS N50,000 ppm), seawater (SW) (TDS 20,000–50,000 ppm), river water (RW) (TDS 500–3000 ppm) and pure water (PW) (TDS b500 ppm) accounting for 69% of the volume of desalinated water produced. is assumed to be the 95th, 90th, 10th and 5th percentiles of brackish water The spatial distribution, size and customer type of desalination facil- technologies respectively. ities (N1000 m3/day) are displayed in Fig. 4. Large numbers of desalina- 2 When brackish water (BW) recovery is unknown but seawater water (SW) tion facilities are located in the United States, China and Australia and recovery is known, the water recovery ratio of brine water (BR), brackish across the regions of Europe, North Africa and the Middle East. Rela- water (BW), river water (RW) and pure water (PW) is assumed to be the 90th, 25th, 10th and 5th percentiles of seawater technologies respectively. tively few desalination facilities are located in South America and 3 The recovery rate of wastewater (WW) for each technology is assumed to be Africa, with existing facilities predominantly designed to produce desa- equal to the recovery rate of brackish water for the same technology. linated water for the industrial sector. Desalination plants globally are Estimation concentrated on and around the coastline, with coastal desalination 1 Other technologies cover a range of different technologies. An estimated 40% water plants also tending to be larger than inland desalination plants. Plants recovery ratio was assigned for highly saline water (above 20,000 ppm) and 60% producing municipal water are located worldwide, but are particularly recovery for brackish and slightly saline water sources (below 20,000 ppm). dominant in the Middle East & North Africa region. Comparatively, E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 1347

Fig. 1. Number of desalination publications by categorisation (total, technical, social, environment, energy & economic).

Fig. 2. Number of publications by type of desalination technology (Reverse Osmosis [RO], Multi-Effect Distillation [MED], Multi-Stage Flash [MSF], Electrodialysis [ED]), emerging technologies (Nanofiltration [NF], Forward Osmosis [FO] and Membrane Distillation [MD]) and other (Humidification-Dehumidification [HDH], Solar Stills [SS] and Vapour Compression [VC]). 1348 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356

Fig. 3. Trends in global desalination by (a) number and capacity of total and operational desalination facilities and (b) operational capacity by desalination technology. there is a far greater proportion of desalination plants producing water accounting for the majority of the global desalination capacity (71%). for non-municipal purposes in North America, Western Europe and East Conversely, very few desalination plants are located in low income Asia and Pacific regions, whereby generation of water for industrial and countries, which contribute a negligible proportion (b0.1%) of the global power applications also command large market shares (Fig. 4). desalination capacity. The number and capacity of desalination plants by geographic re- Whilst almost half of the total number of desalination plants pro- gion, country income level and sectoral use of desalinated water duce water for the industrial sector, the municipal sector is the largest (Table 2) reveal that almost half of the global desalination capacity is lo- user of desalinated water in terms of capacity. 62.3% of desalinated cated in the Middle East and North Africa region (48%), with Saudi water is produced for human consumption (municipal sector), com- Arabia (15.5%), the United Arab Emirates (10.1%) and Kuwait (3.7%) pared to 30.2% for industrial applications. This pattern occurs due to being both the major producers in the region and globally. East Asia the (typically) smaller capacity of industrial desalination facilities, and Pacific and North America regions produce 18.4% and 11.9% of the which average 3712 m3/day, compared to desalination plants producing global desalinated water, primarily due to large capacities in China municipal water that average 12,126 m3/day. Whilst the municipal and (7.5%) and the USA (11.2%). The widespread use of desalination in industrial sectors account for the vast majority of the global desalination Spain (5.7%) accounts for over half of the total desalination in Western capacity, the power (4.8%) and irrigation (1.8%) sectors consume a small Europe (9.2%). The global share in desalination capacity is lower for but significant proportion of produced desalinated water. Southern Asia (3.1%), Eastern Europe and Central Asia (2.4%) and Sub- Of the desalination technologies, RO is by far the most dominant pro- Saharan Africa (1.9%), where desalination is primarily restricted to cess, accounting for 84% of the total number of operational desalination small facilities for private and industrial applications. The majority of plants, producing 69% (65.5 million m3/day) of the total global desali- desalination facilities are located in high income countries (67%), nated water (Fig. 5a). The two major thermal technologies, MSF and

Fig. 4. Global distribution of operational desalination facilities and capacities (N1000 m3/day) by sector user of produced water. E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 1349

Table 2 MSF, SW–MED, RW-RO, WW-RO, BW-ED, RW-ED) are responsible for Number, capacity and global share of operational desalination plants by region, country in- the production of over 90% of the global desalinated water. come level and sector use. Fig. 6 reveals the spatial distribution and size of large Number of desalination Desalination (N10,000 m3/day) desalination plants operating under different plants capacity feedwater-technology combinations. Thermal desalination technologies (million (%) (MED, MSF) operating with sea water as the feedwater type are dominant m3/day) in the Middle East, with the exception of a large number of BW-RO plants Global 15,906 95.37 100 located in inland Saudi Arabia. Outside of this region, very few large ther- Geographic region mal plants exist, with RO being the dominant technology across a range of Middle East and North Africa 4826 45.32 47.5 feedwater types. For example, large desalination plants in Australia oper- East Asia and Pacific 3505 17.52 18.4 ate almost exclusively using RO technology, but with a variety of North America 2341 11.34 11.9 Western Europe 2337 8.75 9.2 feedwater types including SW, BW and WW. RO is also the dominant Latin America and Caribbean 1373 5.46 5.7 technology across the United States, although the vast majority of desali- Southern Asia 655 2.94 3.1 nation plants operate using BW and RW, with only a small number of sea- Eastern Europe and Central Asia 566 2.26 2.4 water plants located in California and Florida. Western Europe, and in Sub-Saharan Africa 303 1.78 1.9 Income level particular Spain, is dominated by RO using a variety of feedwater sources, High 10,684 67.24 70.5 although there is also a significant number of desalination plants operat- Upper middle 3075 19.16 20.1 ing using alternative technologies such as ED and NF. Lastly, SW-RO dom- Lower middle 2056 8.88 9.3 inates desalination in the coastal areas of Asia, although a significant Low 53 0.04 0.0 number of BW- and RW-RO plants are located inland. Sector use Municipal 6055 59.39 62.3 Industry 7757 28.80 30.2 3.3. Brine production Power 1096 4.56 4.8 Irrigation 395 1.69 1.8 The water recovery efficiency of desalination operations depends on Military 412 0.59 0.6 Other 191 0.90 0.4 both the type of desalination technology and the quality of feedwater used, and therefore both of these factors must be considered when quantifying brine production (Xu et al., 2013). Table 3 displays the MED, despite being relatively few in number, produce the majority of water recovery ratios associated with the major feedwater-technology the remaining desalinated water, with market shares of 18% and 7% re- combinations in operation. spectively (Fig. 5a). In total, these three technologies account for 94% of For all technologies, the recovery ratio increases as the feedwater the total desalinated water produced, with plants using NF (3%), ED quality increases (salinity decreases), with BR associated with the low- (2%) and EDI (b1%) technologies producing smaller volume of desali- est water recovery ratios and PW associated with the highest recovery nated water (Fig. 5a). ratios. Feedwater type is a substantial determinant of the recovery In terms of feedwater source, which is indicative of feedwater qual- ratio associated with a particular technology. For example, SW-RO ity, SW desalination accounts for 61% of produced water (Fig. 5b). Desa- operates at a substantially lower recovery ratio (0.42) compared to lination of BW and RW produce the next largest volumes of desalinated BW-RO (0.65) and RW-RO (0.85). Similarly, BW-NF (0.83) is substan- water, with market shares of 21% and 8% respectively (Fig. 5b). In total, tially more efficient than SW-NF (0.69). Individual desalination technol- these three sources account for 90% of the total volume of desalinated ogies are also associated with vastly different recovery ratios. Thermal water produced, with the remainder being produced from WW (6%), technologies (e.g. MSF, MED) are typically associated with much PW (4%) and BR (b1%). lower recovery ratios than membrane technologies (e.g. RO, NF). For ex- Whilst Fig. 5 clearly demonstrates the relative dominance of RO, MSF ample, the recovery ratio of MSF across all feedwater types is approxi- and MED in terms of desalination technology, and SW, BW and RW in mately half that of RO. The water recovery ratio of other membrane terms of feedwater source, the combination of both these factors is im- technologies (NF, ED, EDI, EDR) is substantially higher than RO across portant. Desalination technologies can be considered semi-specialised all feedwater types. in that they operate most efficiently when using particular source Energy requirements, and hence economic costs, vary depending on water types, or that their economic viability is dependent on source feedwater type. For membrane technologies, low salinity feedwater water type, and hence some feedwater-technology combinations are types (e.g. RW) require less applied pressure than high salinity significantly more prevalent than others. feedwater types (e.g. SW) for desalination, causing lower energy con- RO is a process that is economically viable across a range of sumption per unit water produced (Ghaffour et al., 2013). This results feedwater types, and hence the feedwater type used is dependent on in substantially lower investment costs (Ghaffour et al., 2013). How- local availability (Fig. 5). 50% and 27% of the desalinated water that is ever, highly efficient membrane technologies are rarely used for desali- produced from RO desalination plants, accounting for 34% and 19% of nation of highly saline feedwater types, with a total of just 0.01% the global desalination capacity, originates from SW and BW water, re- desalinated water being produced by SW or BR in combination with spectively. RO of RW (7%) and WW (5%) also contributes a significant NF, ED, EDI and EDR. For highly saline feedwater types, RO and thermal proportion of the global desalination capacity. Comparatively, thermal processes (e.g. MSF, MED) dominate. Whilst thermal technologies (par- technologies are used almost exclusively for low quality (highly saline) ticularly MED) are associated with higher energy consumption, the eco- feedwater types. 96% of MSF plants and 80% of MED plants use nomic cost of desalting SW is comparable to RO due to lower feedwater with N20,000 ppm TDS, the vast majority of which use sea investment costs (Ghaffour et al., 2013). water. SW accounts for 99.9% and 92% of the total volume of desalinated Current global brine production stands at 141.5 million m3/day, to- water produced by MSF and MED respectively, representing global mar- taling 51.7 billion m3/year (Table 4). This value is approximately 50% ket shares of 18% and 6%. Conversely, plants operating with ED as the greater than the total volume of desalinated water produced globally. desalination technology typically require water of a higher base quality Global brine production is concentrated in the Middle East and North (lower salinity). 60% and 20% of the desalinated water produced by ED Africa, which produces almost 100 million m3/day of brine, accounting originates as BW and RW respectively, contributing a small but signifi- for 70.3% of global brine production. This value is approximately double cant proportion of the total global volume of desalinated water. In the volume of desalinated water produced, indicating that desalination total, eight feedwater-technology combinations (SW-RO, BW-RO, SW– plants in this region operate at an (very low) average water recovery 1350 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356

Fig. 5. Number and capacity of operational desalination facilities by (a) technology and (b) feedwater type. ratio of 0.25. Comparatively, all other regions produce substantially accounting for 22.2% of the global share. The next three largest pro- lower volumes of brine, with East Asia and Pacific (10.5%), Western ducers of brine are also oil-rich countries, with the UAE, Kuwait and Europe (5.9%) and North America (3.9%) having the next largest shares. Qatar having 20.2%, 6.6% and 5.8% shares in global brine production re- Interestingly, these regions produce a substantially lower volume of spectively. Together, these four nations produce 32% of global desali- brine than the amount of desalinated water they produce, indicating nated water and 55% of the total brine. Comparatively, the USA that recovery ratios are generally high. This is particularly apparent for produces 10.91 million m3/day of desalinated water (11.4% global North America, which produces a substantially lower volume of brine share) but produces just 5.28 million m3/day of brine (3.7% global than it does desalinated water, suggesting that desalination facilities op- share). Upper middle income, lower middle income and low income erate at an average recovery ratio of 0.75. In other geographical regions, countries tend to produce quantities of brine similar to that of their re- brine production is approximately equivalent to desalinated water pro- spective desalination capacities. duction (i.e. RR = 0.5). Water produced for the municipal sector is by far the largest pro- As with desalinated water production, high income countries pro- ducer of both desalinated water and brine, although the quantity of duce the vast majority of global brine (77.9%). It should be noted that brine produced is much greater. This pattern arises primarily due to ‘high income’ includes both countries from both highly developed the vast quantity of desalinated drinking water produced for the Gulf world regions (e.g. North America, Western Europe), whose brine pro- nations, whereby thermal technologies operating with SW dominate. duction tends to be smaller relative to the desalinated water production, Both the industrial and agricultural sectors produce lower quantities and the oil-rich Gulf nations who typically employ thermal desalination of brine than desalinated water, indicating desalinated water for these technologies with low recovery ratios, hence high brine production. For sectors is produced by feedwater-technology combinations with higher example, Saudi Arabia alone produces 31.53 million m3/day brine, water recovery ratios. This is particularly pronounced in the agricultural E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 1351

Fig. 6. Global distribution of large desalination plants by capacity, feedwater type and desalination technology. sector due to the dominance of high-quality (low salinity) feedwater Whilst brine disposal into saline surface water bodies raises some sources used for producing desalinated water for use in agriculture important environmental concerns, this option is extremely economical sector. (Arnal et al., 2005). However, this option is often not available for inland The geographical location of brine production influences the eco- desalination plants, which account for a smaller yet significant propor- nomic and technical viability of different methods of brine disposal tion of the volume of brine being produced. Almost 22 million m3/day (Arnal et al., 2005). Desalination plants located near the shoreline of brine is produced at a distance of N50 km from the nearest coastline often discharge untreated brine directly into saline surface water bodies (Table 5). Despite the large volume of brine produced inland, very few (e.g. oceans, seas) (Arnal et al., 2005). As almost half of brine is pro- economically viable and environmentally sound brine management op- duced within 1 km of the coastline, rising to almost 80% produced tions exist (Arnal et al., 2005). Brine produced inland poses an impor- within 10 km, ocean disposal is assumed to be the dominant brine dis- tant problem for many countries located in all world regions, with 64 posal method worldwide (Table 5). The countries producing large vol- countries producing N10,000 m3/day of brine in inland locations umes of brine (N1 million m3/day) in coastal locations are largely (Fig. 7b). Whereas the volume of brine produced in coastal locations is concentrated in the Middle East and North Africa (e.g. UAE, Saudi largely concentrated in the Middle East, inland brine production is a Arabia) and South-East Asia (China, India), and in the USA and particular issue in other locations such as China (3.82 million m3/day), Australia (Fig. 7a). The volume of brine produced in many of these coun- USA (2.42 million m3/day) and Spain (1.01 million m3/day) (Fig. 7b). tries far exceeds 1 million m3/day, particularly in the Middle East. In this Whilst Fig. 5 considered the production of desalinated water by region, the four largest brine producers (UAE, Saudi Arabia, Qatar, technology and feedwater type separately, Fig. 8a combines these two Kuwait) account for 72.2 million m3/day of the brine that is produced elements, displaying the 6 major feedwater-technology combinations within 10 km of the coastline. by volume of desalinated water produced. As displayed in Fig. 5,ROis

Table 3 Table 4 Recovery ratio of different feedwater-technology combinations producing desalinated Brine production and share of global total by region, income level and sector use. water. Brine production Feedwater type Technology (million m3/day) (%) RO MSF MED NF ED EDI EDR Other Global 141.5 100 Seawater (SW) 0.42 0.22 0.25 0.69 0.86 0.90 0.40 Geographic region Brackish (BW) 0.65 0.33 0.34 0.83 0.90 0.97 0.90 0.60 Middle East & North Africa 99.4 70.3 River (RW) 0.81 0.35 0.86 0.90 0.97 0.96 0.60 East Asia & Pacific 14.9 10.5 Pure (PW)a 0.86 0.35 0.89 0.90 0.97 0.96 0.60 North America 5.6 3.9 Brine (BR) 0.19 0.09 0.12 0.85 0.40 Western Europe 8.4 5.9 Wastewater (WW)b 0.65 0.33 0.34 0.83 0.90 0.97 0.60 Latin America & Caribbean 5.6 3.9 Southern Asia 3.7 2.6 Based on data from: Ahmed et al. (2001), Allison (1993), Almulla et al. (2003), Eastern Europe & Central Asia 2.5 1.8 Bashitialshaaer et al. (2007), Belatoui et al. (2017), Bleninger et al. (2010), Costa and De Sub-Saharan Africa 1.5 1.0 Pinho (2006), DesalData (2018), Efraty and Gal (2012), Fernández-Torquemada et al. Income level (2005), Garcia et al. (2011), Gomez and Cath (2011), Greenlee et al. (2009), Hajbi et al. High 110.2 77.9 (2010), Harvey (2008), Kelkar et al. (2003), Khawaji et al. (2007), Korngold et al. Upper middle 20.7 14.6 (2009), Kurihara et al. (2001), Macedonio and Drioli (2008), Mohamed et al. (2005), Lower middle 10.5 7.4 Mohsen and Gammoh (2010), Pilat (2001), Pearce et al. (2004), Qiu and Davies (2012), Low 0.03 0.0 Qurie et al. (2013), Singh (2009), Stover (2013), Valero and Arbós (2010), Von Gottberg Sector use et al. (2005), Voutchkov (2011), Wilf and Klinko (2001), Xu et al. (2013), Younos Municipal 106.5 75.2 (2005) and Zhou et al. (2015). Industry 27.4 19.3 a PW refers to water of a high base quality (low salinity), but that is desalinated pri- Power stations 5.8 4.1 marily for industrial applications requiring very low salinity water (e.g. food processing, Irrigation 1.1 0.8 pharmaceutical manufacturing). Military 0.5 0.3 b WW refers to reject water from municipal and industrial sources undergoing desali- Other 0.3 0.2 nation in specific WW desalination facilities. 1352 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356

Table 5 operates at very high water recovery ratios, therefore producing Brine production and share of global total by distance to coastline. 6.8 million m3/day of desalinated water (7.1% global share) whilst pro- 3 Distance to coastline Brine production ducing just 1.6 million m /day of brine (1.1% global share). Conversely, whilst SW-MSF desalination plants produce 16.7 million m3/day (million m3/day) (%) (17.6% global share) of desalinated water, brine production totals b 1 km 69.0 48.8 60.1 million m3/day (43% global share). 1km–10 km 43.2 30.5 10 km–50 km 7.3 5.2 N50 km 22.0 15.5 4. Discussion

Owing to recent, rapid developments in desalination research, the the dominant desalination technology, with significant additional con- last comprehensive assessment (Tanaka and Ho, 2011) available in the tributions from MSF and MED technologies (Fig. 8a). However, large academic literature is outdated. This study presents statistical analysis volumes of desalination water are produced by RO from a variety of of the scientific literature covering an array of desalination topics since feedwater sources (SW, BW, RW and WW), whilst the two thermal 1980, addressing a diverse range of social and technical aspects with technologies almost exclusively use SW. The share in brine production respect to publication date, revealing patterns in publishing trends. from each desalination feedwater-technology combination is displayed Our findings suggest that research in desalination has increased expo- in Fig. 8b. The vast majority of brine, 124.5 million m3/day (87.9%), nentially over the last 40 years, coinciding with the more widespread comes from SW desalination plants. Comparatively, brine production recognition of the value of these technologies for water resources man- from desalination plants operating with other feedwater types is agement. Research has particularly considered the technological aspects much smaller, with BW, RW and WW plants producing 10.23 (7.2%), of desalination, with the vast number of publications addressing RO and 1.80 (1.3%) and 3.57 (2.5%) million m3/day, respectively. Individually, novel (‘emerging’) techniques that can produce desalinated water at SW-MSF accounts for the largest volume of brine production (43%), lower economic costs and with less negative environmental implica- with SW-RO (31%), SW-MED (12%) and BW-RO (7%) accounting for tions (Arnal et al., 2005). Developments in novel desalination tech- the vast majority in the remainder of the global share (Fig. 8b). niques, membrane materials and modules dominate the technological Clear discrepancies exist when comparing the volume of desalinated literature (Greenlee et al., 2009). Publications addressing economic, en- water produced to the volume of brine water produced by different vironmental and socio-political aspects of desalination are rapidly feedwater-technology combinations (Fig. 8c). These differences are di- increasing with expansions in desalination. In particular, the environ- rectly related to the different water recovery ratios associated with de- mental impacts associated with hypersaline brine discharges from desa- salination plants operating with different feedwater-technology lination plants have received increased attention in recent research, combinations. The greater the volumetric processing efficiency of the coinciding with water scarcity intensification and the resultant expan- desalination process, the smaller the proportion of brine produced rela- sion in desalination operations in the USA, Europe and Australia tive to the volume of desalinated water produced. For example, RW-RO (Roberts et al., 2010).

Fig. 7. Volume of brine produced per country at a distance of a) b10 km and b) N50 km from the coastline. E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 1353

Fig. 8. Major desalination feedwater-technology combinations by (a) global share in the desalinated water production (%), (b) global share in brine production (%) and (c) total desalination capacity and volume of brine produced (million m3/day).

With regards to desalination in practice, the current state-of-the-art is an essential technology in the Middle East and for small island in the global desalination situation in existing academic literature is ei- nations which typically lack renewable water resources. ther a) severely outdated or b) derived from incomplete sources. This Whilst there are demonstrated benefits from desalinated water, study uses the largest and the most complete desalination dataset avail- there are concerns related to the volume and salinity of brine produced able (DesalData, 2018) to comprehensively analyse the global state of as a waste of desalination process. It poses some of the biggest con- desalination with respect to desalination geographical distribution, sec- straints to more widespread development of desalination operations, tor use, technology and feedwater type. Our findings demonstrate that in addition to representing a significant proportion of the economic the global desalination capacity far exceeds values frequently cited in costs of the process (5–33%) (Ahmed et al., 2001). Therefore, quantify- the literature, due to both the vast expansion in desalination operations ing the volume of brine produced by desalination plants operating that have taken place across the globe in the last decade and the cover- with different feedwater types and technologies is essential for consid- age of the dataset used. The major uncertainty related to these results is ering the potential environmental and economic costs associated with due to the completeness of the desalination database, which was desalination. minimised by using a very high-quality dataset (DesalData, 2018). The volume of brine being produced from desalination plants glob- Accurate quantifications of the volume of desalinated water pro- ally is largely unknown, with the only estimates available in the litera- duced for human use at different spatial scales is associated with a ture assuming that the volume of brine is equivalent to the volume of range of management implications. For example, in order to more rep- desalinated water produced (Liu et al., 2016; Akinaga et al., 2018), re- resentatively assess the degree of water scarcity, unconventional water gardless of the feedwater type or desalination technology. Our study resources and management practices must be included (Vanham et al., considers the influence of both feedwater type and desalination tech- 2018; Jones and van Vliet, 2018). The exclusion of desalination from nology on the water recovery ratio, deriving values for the vast majority quantifications of water scarcity is identified as a major shortcoming of feedwater-technology combinations producing desalinated water. of SDG 6.4.2 (Vanham et al., 2018), and hence accurate data on the vol- This information is applied with respect to the treatment capacity of in- ume of desalinated water produced is important for assessing the actual dividual desalination plant to determine the volume of brine produc- status of water availability. This is particularly important as water scar- tion, thus representing a first comprehensive attempt to accurately city has been identified as a key challenge, which is expected to inten- quantify brine production. sify in the future (Richter et al., 2013). Our findings also indicate that the volume of brine produced far ex- The importance of desalination for alleviating water scarcity ceeds the volume of desalinated water produced (by ~50%), and hence and safeguarding water resources for human use should not be that the current quantifications of volume of brine produced are gross underestimated. Based on FAO AQUASTAT water withdrawal data underestimations. However, the uncertainties associated with our (http://www.fao.org/nr/water/aquastat/water_res/index.stm) method should be considered. In our brine assessment methodology, and desalination capacity data from DesalData (2018),eightcoun- we assigned recovery ratios based solely on the feedwater type and de- tries produce more desalination water than they withdraw for salination technology producing desalinated water at each plant, with human use (The Maldives, Singapore, Qatar, Malta, Antigua and no consideration of local conditions in these plants. Evidence suggests Barbuda, Kuwait, The Bahamas and Bahrain). A further six coun- that site-specific local conditions may also influence a desalination tries meet over 50% of their water withdrawals through desalina- plants recovery ratio (Xu et al., 2013). For example, the effect of varia- tion (Equatorial Guinea, UAE, Seychelles, Cape Verde, Oman and tions in feedwater salinity within each ‘feedwater type’ categorisation Barbados). As demonstrated through these countries, desalination (e.g. seawater) on the recovery ratio is overlooked as a result of using 1354 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 the average. Other factors that may influence the specific recovery of bottom ponding (Roberts, 2015). Techniques such as bipolar membrane each individual desalination plant include specificplantdesign(e.g. electrodialysis (BMED) can convert brine into acid and base products for type of membrane used in desalination process), product water quality reuse, such as NaOH and HCl (Ibáñez et al., 2013; Morillo et al., 2014). requirements (e.g. salinity), energy source and brine disposal method- Metal recovery from brine offers a valuable source of many scarce ology. Furthermore, whilst desalination plant recovery ratios are avail- metals (e.g. uranium), whilst potentially reducing environmental im- able in the literature, the number of values used for determining pacts associated with mining (Morillo et al., 2014; Loganathan et al., recovery ratios for some feedwater-technology combinations was low. 2017). The high economic costs and energy demands of brine treatment For some feedwater-technology combinations, no values were found and mineral recovery methods remain a significant barrier to more in the literature, and therefore a number of assumptions and estima- widespread application (Kaplan et al., 2017). Comprehensive reviews tions had to be made (Table 1). This uncertainty was minimised as re- of the recent techniques, technologies and innovations in brine manage- covery ratios were found in the literature for all major feedwater- ment are provided by Morillo et al. (2014) and Giwa et al. (2017). technology combinations, capturing the vast majority of the total desa- Other potential economic opportunities associated with brine pro- lination capacity (N80%). duction have also sparked a wave in innovation in brine management With increasing water demands coupled with water scarcity intensi- that seeks to turn an environmental problem into an economic opportu- fication, desalination is expected to expand rapidly in the future. The ex- nity (Sánchez et al., 2015). For example, Blackwell et al. (2005) identi- pected expansion in desalination capacity will be commensurate with fied sequential biological concentration (SBC) of saline drainage an increase in the volume of brine produced. Management of the reject streams creating a number of financial opportunities, whilst concentrat- brine is the still a major problem of desalination (Roberts et al., 2010; ing the waste stream into a manageable volume. Qadir et al. (2015) sug- Elimelech and Phillip, 2011; Mezher et al., 2011; Wenten, 2016), con- gested that integrating agriculture and aquaculture systems based on taining both elevated salinity (relative to feedwater type) and chemicals the SBC system using saline drainage water sequentially has the poten- used during pre- and post-treatment phases in the desalination opera- tial for commercial, social and environmental gains. Reject brine has tion (Wenten et al., 2016). Traditionally, a variety of brine disposal been used for aquaculture, with increases in fish biomass of 300% methods have been used, including direct discharge into oceans, surface achieved (ICBA, 2018). Reject brine has also been successfully used for water or sewers, deep well injection and brine evaporation ponds Spirulina cultivation and the irrigation of halophytic forage shrubs and (Morillo et al., 2014). The geographical location at which brine is pro- crops although this method was unable to prevent progressive land duced influences the brine disposal method – desalination plants lo- salinisation (Sánchez et al., 2015). cated near to large surface saline water bodies (ocean, seas) often Aside from treating or using reject brine, a method to reduce the discharge untreated waste brine directly into these water bodies volume of brine produced is to improve the water recovery ratio of de- (Arnal et al., 2005). Conversely, desalination plants located inland may salination plants. Desalination plants operating with a high RR are not have a surface water discharge option available, and hence alterna- favourable in that they both maximise the use of (often scarce) water tive brine disposal methods are required, of which there are few eco- resources as in the case of river and brackish water desalination plants nomically viable options (Arnal et al., 2005; Brady et al., 2005; Morillo and create a lower volume of concentrate for disposal (Harvey, 2008), et al., 2014). reducing the economic costs associated with brine disposal. High recov- Whilst the majority of brine is produced near to the coastline ery rates can also reduce the cost of pre-treatment prior to desalination (Fig. 7a), with almost 80% of brine produced within 10 km (Table 5), a and post-treatment of brine (Lachish, 2002). However, attaining higher substantial volume of brine is produced in geographic locations where RRs generally increases energy demands and hence treatment costs surface water discharge is likely not possible (Fig. 7b, Table 5). In addi- (Lachish, 2002), increasing greenhouse gas emissions if the desalination tion, there are a variety of environmental concerns associated with the plant is powered by fossil fuels (Martin-Gorriz et al., 2014; Darre and discharge of hypersaline brine into surface water bodies (Einav et al., Toor, 2018). Whilst the reduced volume of brine associated with higher 2002; Roberts et al., 2010; Palomar and Losada, 2011). Major concerns RRs might have positive environmental implications, the brine salt con- are related to the ecological effects associated with physio-chemical al- centration will be increased (Ahmed et al., 2001) which could poten- terations (e.g. increased salinity) to seawater around brine discharge tially pose harmful risks to the aquatic environment following disposal outlets and the discharge of toxic chemicals used in water pre- (Bashitialshaaer et al., 2009). Determining the optimal recovery ratio treatment or as anti-scalants and anti-foulants in the desalination pro- for desalination plants is therefore an economic, environmental and cess (Einav et al., 2002; Roberts et al., 2010; Ketsetzi et al., 2008). technical challenge, requiring consideration on a site-by-site basis. When continually discharged to surface waters, these factors pose risks to ocean life and marine ecosystems (Gacia et al., 2007; Palomar 5. Conclusions & outlook and Losada, 2011; Meneses et al., 2010). The high salinity of brine causes elevated density in comparison to the salinity of the receiving waters, Against the backdrop of increasing global water scarcity, desalinated which can form “brine underflows” that deplete dissolved oxygen water is increasingly becoming a viable option to narrow the water (DO) in the receiving waters. High salinity and reduced DO levels can demand-supply gap, particularly in addressing domestic and municipal have profound impacts on benthic organisms, which can translate into needs. Desalinated water can substantially extend the volume of high- ecological effects observable throughout the food chain (Rabinowitz, quality water supplies available for human use. A steady and assured 2016; Frank et al., 2017). A combination of these factors necessitates supply of high-quality water is crucially important in an era when the the development of new brine management strategies that are both world at large is embarking on the Sustainable Development Agenda economically feasible and environmentally sound. to ensure access to safe water for all by 2030, and for the achievement Recent efforts have focused on ways to treat or use brine in order to of SDG 6 to safeguard water supplies for current and future generations. minimise or eliminate the negative environmental impacts associated In addition to SDG 6, a variety of other SDGs are inextricably linked with with brine disposal (Morillo et al., 2014; Wenten et al., 2017) and/or water resources management, such as SDG 2 aiming at zero hunger, to partially or fully offset the economic costs associated with brine dis- SDG 3 ensuring healthy lives, SDG 8 promoting sustainable economic posal (Kesieme et al., 2013; Morillo et al., 2014). These efforts cover a growth, SDG 11 making cities and human settlements inclusive, and range of techniques with variable levels of complexity and cost. For ex- SDG 13 combating climate change. These SDGs have water-related tar- ample, mixing brine with alternative water sources of a lower salinity gets that must be achieved before their ultimate realisation is possible. (e.g. treated wastewater, power-plant cooling water) can reduce brine Although desalination can provide an unlimited, climate-independent salinity by dilution (Giwa et al., 2017). Pressurised dispersion nozzles and steady supply of high-quality water, there are specific challenges to can promote mixing of brine waters with receiving waters, restricting harness the vast potential of desalinated water, such as relatively high E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356 1355 economic costs and a variety of environmental concerns. A major envi- Bashitialshaaer, R., Persson, K.M., Larson, M., Ergil, M., 2007. Impact of brine disposal from EMU desalination plant on seawater composition. IDA World Congress Maspalomas, ronmental concern arises from the large volume of brine produced in Gran Canaria–Spain October, pp. 21–26. the desalination process that requires management. Brine management Bashitialshaaer, R., Flyborg, L., Persson, K.M., 2009. Environmental assessment of brine is both economically expensive and technically difficult, and hence most discharge including wastewater collection in the Arabian gulf. Proceedings 12th IDA World Congress & Water Reuse, Nov. 7–12. desalination plants discharge untreated brine directly into the environ- Belatoui, A., Bouabessalam, H., Rouane Hacene, O., De-la-Ossa-Carretero, J.A., Martinez- ment. Addressing these challenges, research studies have demonstrated Garcia, E., Sanchez-Liza, J.L., 2017. Environmental effects of brine discharge from that there are economic opportunities associated with brine, such as com- two desalination plants in Algeria (South Western Mediterranean). Desalin. Water – mercial salt and metal recovery and use of brine in fish and halophyte Treat. 76, 311 318. Blackwell, J., Jayawardane, N., Biswas, T.K., Christen, E.W., 2005. Evaluation of a sequential production systems. There is a need to translate such research to convert biological concentration system in natural resource management of a saline irrigated an environmental problem into an economic opportunity. This is particu- area. Aust. J. Water Resour. 9, 169–176. Bleninger, T., Niepelt, A., Jirka, G., 2010. Desalination plant discharge calculator. Desalin. larly important in countries producing large volumes of brine with rela- – – fi Water Treat. 13 (1 3), 156 173. tively low ef ciencies, such as Saudi Arabia, UAE, Kuwait and Qatar. Brady, P.V., Kottenstette, R.J., Mayer, T.M., Hightower, M.M., 2005. Inland desalination: Although smaller amounts of desalinated water are used for the challenges and research needs. J. Contemp. Water Res. Educ. 132 (1), 46–51. power and irrigation sectors, water is desalinated primarily for mu- Caldera, U., Breyer, C., 2017. Learning curve for seawater reverse osmosis desalination plants: capital cost trend of the past, present, and future. Water Resour. Res. 53 nicipal and industrial purposes. In this regard, desalinated water (12), 10523–10538. provides a safe and sustainable source of good-quality water for do- Chen, T., Wang, Q., Qin, Y., Chen, X., Yang, X., Lou, W., Zhou, M., He, G., Lu, K., 2015. Knowl- mestic purposes. Such potable water supplies are critically impor- edge, attitudes and practice of desalinated water among professionals in health and water departments in Shengsi, China: a qualitative study. PLoS One 10 (4), e0118360. tant in water scarce areas where water quality deterioration is also Costa, A.R., De Pinho, M.N., 2006. Performance and cost estimation of nanofiltration for on the increase. The use of desalinated water in producing high- surface water treatment in drinking water production. Desalination 196 (1–3), 55–65. value crops and crop commodities would be another avenue whilst Damania, R., Desbureaux, S., Hyland, M., Islam, A., Moore, S., Rodella, A.S., Russ, J., Zaveri, E., 2017. Uncharted Waters: The New Economics of Water Scarcity and Variability. considering expansion of desalinated water to other sectors (Silber World Bank, Washington, DC (© World Bank. https://openknowledge.worldbank. et al., 2015). org/handle/10986/28096 License: CC BY 3.0 IGO). Due primarily to the relatively high economic costs, desalination is Darre, N.C., Toor, G.S., 2018. Desalination of water: a review. Curr. Pollut. Rep. 1–8. currently concentrated in high income and developed countries. There DesalData, 2018. GWI/DESALDATA. Available at:. https://www.desaldata.com/,Accessed date: 7 May 2018. is a need to make desalination technologies more affordable and extend Djuma, H., Bruggeman, A., Eliades, M., Lange, M.A., 2016. Non-conventional water re- them to low income and lower middle income countries, increasing the sources research in semi-arid countries of the Middle East. Desalin. Water Treat. 57 – viability of desalination for addressing SDG 6 in areas that develop- (5), 2290 2303. Efraty, A., Gal, Z., 2012. Closed circuit desalination series no 7: retrofit design for improved ments have previously been limited by high economic costs. To do performance of conventional BWRO system. Desalin. Water Treat. 41 (1–3), 301–307. this, technological refinement for low environmental impacts and eco- Einav, R., Harussi, K., Perry, D., 2002. The footprint of the desalination processes on the en- nomic costs, along with innovative financial mechanisms to support vironment. Desalination 152 (1–3), 141–154. Elimelech, M., Phillip, W.A., 2011. The future of seawater desalination: energy, technol- the sustainability of desalination schemes, will likely be required. The ogy, and the environment. Science 333 (6043), 712–717. expansion pattern and economics of desalination facilities in recent de- Fernández-Torquemada, Y., Sánchez-Lizaso, J.L., González-Correa, J.M., 2005. Preliminary cades suggest a positive and promising outlook for expansion in desali- results of the monitoring of the brine discharge produced by the SWRO desalination plant of Alicante (SE Spain). Desalination 182 (1–3), 395–402. nation facilities around the world. Frank, H., Rahav, E., Bar-Zeev, E., 2017. Short-term effects of SWRO desalination brine on benthic heterotrophic microbial communities. Desalination 417, 52–59. Acknowledgements Gacia, E., Invers, O., Manzanera, M., Ballesteros, E., Romero, J., 2007. Impact of the brine from a desalination plant on a shallow seagrass (Posidonia oceanica) meadow. Estuar. Coast. Shelf Sci. 72 (4), 579–590. This work is part of the UNU-INWEH's project on Unconventional Garcia, C., Molina, F., Zarzo, D., 2011. 7 year operation of a BWRO plant with raw water from a Water Resources. UNU-INWEH is supported by the Government of coastalaquiferforagriculturalirrigation. Desalin. Water Treat. 31 (1–3), 331–338. Canada through Global Affairs Canada. Michelle van Vliet was finan- Ghaffour, N., Missimer, T.M., Amy, G.L., 2013. Technical review and evaluation of the eco- nomics of water desalination: current and future challenges for better water supply cially supported by a Veni-grant (project no. 863.14.008) of NWO Earth sustainability. Desalination 309, 197–207. and Life Sciences (ALW). Giwa, A., Dufour, V., Al Marzooqi, F., Al Kaabi, M., Hasan, S.W., 2017. Brine management methods: recent innovations and current status. Desalination 407, 1–23. fl Gomez, J.D., Cath, T.Y., 2011. Assessment of Osmotic Mechanisms Pairing Desalination Con icting interests Concentrate and Wastewater Treatment. CH2M HILL. Greenlee, L.F., Lawler, D.F., Freeman, B.D., Marrot, B., Moulin, P., 2009. Reverse osmosis de- The authors declare no conflict of interest. salination: water sources, technology, and today's challenges. Water Res. 43 (9), 2317–2348. Gude, V.G., 2017. Desalination and water reuse to address global water scarcity. Rev. En- Appendix A. Supplementary information viron. Sci. Biotechnol. 16 (4), 591–609. Hajbi, F., Hammi, H., M'nif, A., 2010. Reuse of RO desalination plant reject brine. J. Phase Equilib. Diffus. 31 (4), 341–347. Supplementary data to this article can be found online at https://doi. Harvey, W., 2008. Desalination efficiency: energy, water and other resources. Available at: org/10.1016/j.scitotenv.2018.12.076. . Suez Water Technologies & Solutionshttps://www.suezwatertechnologies.com/ kcpguest/salesedge/...Cust/.../TP1154EN.pdf, Accessed date: 26 July 2018. Ibáñez, R., Pérez-González, A., Gómez, P., Urtiaga, A.M., Ortiz, I., 2013. Acid and base recov- ery from softened reverse osmosis (RO) brines. Experimental assessment using References model concentrates. Desalination 309, 165–170. International Center for Biosaline Agriculture [ICBA], 2018. ICBA scientists manage to in- Ahmed, M., Shayya, W.H., Hoey, D., Al-Handaly, J., 2001. Brine disposal from reverse os- crease fish biomass by 300% using reject brine. Available at:. https://www.biosaline. mosis desalination plants in Oman and the United Arab Emirates. Desalination 133 org/news/2018-06-07-6506, Accessed date: 8 February 2018. (2), 135–147. Jones, E., van Vliet, M.T.H., 2018. Drought impacts on river salinity in the southern US: im- Akinaga, T., Generalis, S.C., Paton, C., Igobo, O.N., Davies, P.A., 2018. Brine utilisation for plications for water scarcity. Sci. Total Environ. 644, 844–853. https://doi.org/ cooling and salt production in wind-driven seawater greenhouses: design and 10.1016/j.scitotenv.2018.06.373. modelling. Desalination 426, 135–154. Kaplan, R., Mamrosh, D., Salih, H.H., Dastgheib, S.A., 2017. Assessment of desalination

Allison, R.P., 1993. High water recovery with electrodialysis reversal. Proceedings technologies for treatment of a highly saline brine from a potential CO2 storage site. American Water Works Assoc. Membrane Conference, pp. 1–4. Desalination 404, 87–101. Almulla, A., Eid, M., Côté, P., Coburn, J., 2003. Developments in high recovery brackish Kelkar, P.S., Joshi, V.A., Ansari, M.H., Manivel, U., 2003. Performance evaluation of reverse water desalination plants as part of the solution to water quantity problems. Desali- osmosis desalination plants for rural water supply in a developing country–a case nation 153 (1–3), 237–243. study. Environ. Monit. Assess. 89 (3), 243–261. Arnal, J.M., Sancho, M., Iborra, I., Gozálvez, J.M., Santafé, A., Lora, J., 2005. Concentration of Kesieme, U.K., Milne, N., Aral, H., Cheng, C.Y., Duke, M., 2013. Economic analysis of desa- brines from RO desalination plants by natural evaporation. Desalination 182 (1–3), lination technologies in the context of carbon pricing, and opportunities for mem- 435–439. brane distillation. Desalination 323, 66–74. 1356 E. Jones et al. / Science of the Total Environment 657 (2019) 1343–1356

Ketsetzi, A., Stathoulopoulou, A., Demadis, K.D., 2008. Being “green” in chemical water wastewater treatment plant using epuvalisation system. Int. J. Mol. Sci. 14 (7), treatment technologies: issues, challenges and developments. Desalination 223 13808–13825. (1–3), 487–493. Rabinowitz, O., 2016. Nuclear energy and desalination in Israel. Bull. At. Sci. 72 (1), 32–38. Khawaji, A.D., Kutubkhanah, I.K., Wie, J.M., 2007. A 13.3 MGD seawater RO desalination Richter, B.D., Abell, D., Bacha, E., Brauman, K., Calos, S., Cohn, A., Disla, C., Friedlander plant for Yanbu Industrial City. Desalination 203 (1–3), 176–188. O'Brien, S., Hodges, D., Kaiser, S., Loughran, M., Mestre, C., Reardon, M., Siegfried, E., Korngold, E., Aronov, L., Daltrophe, N., 2009. Electrodialysis of brine solutions discharged 2013. Tapped out: how can cities secure their water future? Water Policy 15 (3), from an RO plant. Desalination 242 (1–3), 215–227. 335–363. Kurihara, M., Yamamura, H., Nakanishi, T., Jinno, S., 2001. Operation and reliability of very Roberts, P.J., 2015. Near field flow dynamics of concentrate discharges and diffuser design. high-recovery seawater desalination technologies by brine conversion two-stage RO Intakes and Outfalls for Seawater Reverse-osmosis Desalination Facilities. Springer, desalination system. Desalination 138 (1–3), 191–199. Cham, pp. 369–396. Lachish, U., 2002. Optimizing the efficiency of reverse osmosis seawater. Desalination Roberts, D.A., Johnston, E.L., Knott, N.A., 2010. Impacts of desalination plant discharges on https://doi.org/10.13140/RG.2.2.12099.20003. the marine environment: a critical review of published studies. Water Res. 44 (18), Liu, J., Yuan, J., Ji, Z., Wang, B., Hao, Y., Guo, X., 2016. Concentrating brine from seawater 5117–5128. desalination process by nanofiltration–electrodialysis integrated membrane technol- Sánchez, A.S., Nogueira, I.B.R., Kalid, R.A., 2015. Uses of the reject brine from inland desa- ogy. Desalination 390, 53–61. lination for fish farming, Spirulina cultivation, and irrigation of forage shrub and Loganathan, P., Naidu, G., Vigneswaran, S., 2017. Mining valuable minerals from seawater: crops. Desalination 364, 96–107. a critical review. Environ. Sci.: Water Res. Technol. 3 (1), 37–53. Schewe, J., Heinke, J., Gerten, D., Haddeland, I., Arnell, N.W., Clark, D.B., Dankers, R., Eisner, Lucier, K., Qadir, M., 2018. Gender and community mainstreaming in fog water collection S., Fekete, B.M., Colon-Gonzalez, F.J., Gosling, S.N., Kim, H., Liu, X., Masaki, Y., systems. Water 10 (10), 1472. Portmann, F., Satoh, Y., Stacke, T., Tang, Q., Wada, Y., Wisser, D., Albrecht, T., Frieler, Macedonio, F., Drioli, E., 2008. Pressure-driven membrane operations and membrane dis- K., Piontek, F., Warszawski, L., Kabat, P., 2014. Multimodel assessment of water scar- tillation technology integration for water purification. Desalination 223 (1–3), city under climate change. Proc. Natl. Acad. Sci. 111 (9), 3245–3250. 396–409. Silber, A., Israeli, Y., Elingold, I., Levi, M., Levkovitch, I., Russo, D., Assouline, S., 2015. Irri- March, H., Saurí, D., Rico-Amorós, A.M., 2014. The end of scarcity? Water desalination as gation with desalinated water: a step toward increasing water saving and crop yields. the new cornucopia for Mediterranean Spain. J. Hydrol. 519, 2642–2651. Water Resour. Res. 51 (1), 450–464. Martinez-Alvarez, V., Martin-Gorriz, B., Soto-García, M., 2016. Seawater desalination for Singh, R., 2009. Analysis of high recovery brackish water desalination processes using fuel crop irrigation—a review of current experiences and revealed key issues. Desalination cells. Sep. Sci. Technol. 44 (3), 585–598. 381, 58–70. Sood, A., Smakhtin, V., 2014. Can desalinization and clean energy combined help to alle- Martin-Gorriz, B., Soto-García, M., Martínez-Alvarez, V., 2014. Energy and greenhouse-gas viate global water scarcity? J. Am. Water Resour. Assoc. https://doi.org/10.1111/ emissions in irrigated agriculture of SE (southeast) Spain. Effects of alternative water jawr.12174. supply scenarios. Energy 77, 478–488. Stover, R.L., 2013. Industrial and brackish water treatment with closed circuit reverse os- Mekonnen, M.M., Hoekstra, A.Y., 2016. Four billion people facing severe water scarcity. mosis. Desalin. Water Treat. 51 (4–6), 1124–1130. Sci. Adv. 2 (2), e1500323. Subramani, A., Jacangelo, J.G., 2015. Emerging desalination technologies for water treat- Meneses, M., Pasqualino, J.C., Céspedes-Sánchez, R., Castells, F., 2010. Alternatives for re- ment: a critical review. Water Res. 75, 164–187. ducing the environmental impact of the main residue from a desalination plant. Tanaka, H., Ho, Y.S., 2011. Global trends and performances of desalination research. J. Ind. Ecol. 14 (3), 512–527. Desalin. Water Treat. 25 (1–3), 1–12. Mezher, T., Fath, H., Abbas, Z., Khaled, A., 2011. Techno-economic assessment and envi- Valero, F., Arbós, R., 2010. Desalination of brackish river water using Electrodialysis Rever- ronmental impacts of desalination technologies. Desalination 266 (1–3), 263–273. sal (EDR): control of the THMs formation in the Barcelona (NE Spain) area. Desalina- Mohamed, A.M.O., Maraqa, M., Al Handhaly, J., 2005. Impact of land disposal of reject tion 253 (1–3), 170–174. brine from desalination plants on soil and groundwater. Desalination 182 (1–3), van Vliet, M.T., Flörke, M., Wada, Y., 2017. Quality matters for water scarcity. Nat. Geosci. 411–433. 10 (11), 800. Mohsen, M.S., Gammoh, S., 2010. Performance evaluation of reverse osmosis desalination Vanham, D., Hoekstra, A.Y., Wada, Y., Bouraoui, F., de Roo, A., Mekonnen, M.M., van de plant: a case study of Wadi Ma'in, Zara and Mujib plant. Desalin. Water Treat. 14 Bund, W.J., Batelaan, O., Pavelic, P., Bastiaanssen, W.G.M., Kummu, M., Rockstrom, (1–3), 265–272. K., Liu, J., Bisselink, B., Ronco, P., Pistocchi, A., Bidoglio, G., 2018. Physical water scar- Morillo, J., Usero, J., Rosado, D., El Bakouri, H., Riaza, A., Bernaola, F.J., 2014. Comparative city metrics for monitoring progress towards SDG target 6.4: an evaluation of indica- study of brine management technologies for desalination plants. Desalination 336, tor 6.4. 2 “level of water stress”. Sci. Total Environ. 613, 218–232. 32–49. Von Gottberg, A., Pang, A., Talavera, J.L., 2005. Optimizing water recovery and energy con- Palomar, P., Losada, I.J., 2011. Impacts of brine discharge on the marine environment. sumption for seawater RO systems. GE Water Process Technol., TP1021EN. 510, Modelling as a Predictive Tool, Desalination Michael Schorr. IntechOpen https:// pp. 1–5. doi.org/10.5772/14880 Available from:. https://www.intechopen.com/books/desali- Voutchkov, N., 2011. Overview of seawater concentrate disposal alternatives. Desalina- nation-trends-and-technologies/impacts-of-brine-discharge-on-the-marine-envi- tion 273 (1), 205–219. ronment-modelling-as-a-predictive-tool. Wenten, I.G., 2016. Reverse osmosis applications: prospect and challenges. Desalination Pearce, G., Talo, S., Chida, K., Basha, A., Gulamhusein, A., 2004. Pretreatment options for 391, 112–125. large scale SWRO plants: case studies of OF trials at Kindasa, Saudi Arabia, and con- Wenten, I.G., Khoiruddin, K., Aryanti, P.T.P., Hakim, A.N., 2016. Scale-up Strategies for ventional pretreatment in Spain. Desalination 167, 175–189. Membrane-Based Desalination Processes: A Review. J. Membr. Sci. Res. 2 (2), 42–58. Pilat, B., 2001. Practice of water desalination by electrodialysis. Desalination 139 (1–3), Wenten, I.G., Ariono, D., Purwasasmita, M., Khoirudin, 2017. Integrated processes for de- 385–392. salination and salt production: a mini-review. AIP Conference Proceedings. vol. 1818 Pinto, F.S., Marques, R.C., 2017. Desalination projects economic feasibility: a standardiza- (No. 1). AIP Publishing, p. 020065. tion of cost determinants. Renew. Sust. Energ. Rev. 78, 904–915. Wilf, M., Klinko, K., 2001. Optimization of seawater RO systems design. Desalination 138 Qadir, M., 2018. Addressing trade-offs to promote safely managed wastewater in devel- (1–3), 299–306. oping countries. Water Econ. Pol. 4 (2), 1871002 (10 pages). Xu, P., Cath, T.Y., Robertson, A.P., Reinhard, M., Leckie, J.O., Drewes, J.E., 2013. Critical re- Qadir, M., Noble, A.D., Karajeh, F., George, B., 2015. Potential business opportunities from view of desalination concentrate management, treatment and beneficial use. Environ. saline water and salt-affected land resources. Resource Recovery and Reuse Series 5 Eng. Sci. 30 (8), 502–514. vol. 5. International Water Management Institute (IWMI). CGIAR Research Program Younos, T., 2005. Environmental issues of desalination. J. Contemp. Water Res. Educ. 132 on Water, Land and Ecosystems (WLE). https://doi.org/10.5337/2015.206 (29 pp.). (1), 11–18. Qadir, M., Jiménez, G.C., Farnum, R.L., Dodson, L.L., Smakhtin, V., 2018. Fog water collec- Zhou, D., Zhu, L., Fu, Y., Zhu, M., Xue, L., 2015. Development of lower cost seawater desa- tion:challengesbeyondtechnology.Water10(4),372. lination processes using nanofiltration technologies—a review. Desalination 376, Qiu, T., Davies, P.A., 2012. Comparison of configurations for high-recovery inland desali- 109–116. nation systems. Water 4 (3), 690–706. Qurie, M., Abbadi, J., Scrano, L., Mecca, G., Bufo, S.A., Khamis, M., Karaman, R., 2013. Inland treatment of the brine generated from reverse osmosis advanced membrane